preg_replace keep the space if start with a-z and end with a-z

Question

I want to replace any space in $string with an underline

but NOT if the space is surrounded by [a-z]

$string = '@@ @@ English something else';

I tried this but it replaces "@@ @@" to "@_@"

$string = preg_replace('/[^a-z]\s[^a-z]/ui', '_' , $string);

I want the output to look like this

@@_@@_English something else

What language is this? 英文中文 – ihodonald Dec 03 '17 at 02:59 — ihodonald, Dec 03 '17 at 02:59

mickmackusa · Accepted Answer · 2017-12-04T06:59:08.730

2

I don't think you need to include the u unicode flag for this task.

Simply disqualify the undesirable spaces with (*SKIP)(*FAIL) then replace all qualifying spaces.

Code: (Demo) (Pattern Demo)

$string = '英文 中文 English something else';
echo preg_replace('/[a-z] [a-z](*SKIP)(*FAIL)| /i','_',$string);

Output:

英文_中文_English something else

Pattern Explanation:

[a-z] [a-z](*SKIP)(*FAIL) <-- this means match and discard the letter-space-letter occurrences
| <-- this mean "OR"
<-- this will match every non-discarded space in the string (only these will be replaced

Two alternatives with lookarounds will also work (and faster): (Demo)

/(?<![a-z]) | (?![a-z])/i

Fastest of all will be the \B (negated version of the word boundary metacharacter): (Demo)

/\B | \B/

edited Dec 04 '17 at 06:59

answered Dec 03 '17 at 03:28

mickmackusa

37,596
11
75
105

must follow a | and the space ? never know this (*SKIP)(*FAIL) very useful – kenken9999 Dec 03 '17 at 03:55
I've added a little more explanation about the pattern. (and, yes, I agree `(*SKIP)(*FAIL)` is VERY useful.) – mickmackusa Dec 03 '17 at 04:00
1

@kenken9999 I just added a couple of new (faster) patterns that will provide the same effect on your sample input. Just thought I'd let you. – mickmackusa Dec 04 '17 at 07:00

score 0 · Answer 2 · answered Dec 03 '17 at 02:58

0

I'm not sure what language that is but I found this: Use regular expression to match ANY Chinese character in utf-8 encoding

Your regular expression needs to contain (all) Chinese characters, which can be expressed as:

\p{Han}

Here is PHP's documentation on Unicode characters and a list of supported scripts: http://php.net/manual/en/regexp.reference.unicode.php

answered Dec 03 '17 at 02:58

ihodonald

649
1
11
26

oh sorry, this is chinese, i edited the question, actually is the same case – kenken9999 Dec 03 '17 at 03:17
This post offers three fractional pieces of advice; two of which are just links. This post would be better positioned as a comment under the question as it doesn't attempt to answer the question. – mickmackusa Dec 04 '17 at 06:48
Just a suggestion... ;] – ihodonald Dec 04 '17 at 20:31

score 0 · Answer 3 · answered Dec 04 '17 at 10:38

sorry i don't know how to use code in comment, so just add a reply

tested the 2nd and 3rd pattern also work

i understand the 2nd one, but the 3rd one a bit hard for me LOL

i need to go check https://www.regular-expressions.info/

$string = '英文名 中文名 Peter Wong，John Tam，Sam Hu';
$string = preg_replace('/(?<![a-z]) | (?![a-z])/i', '，', $string);

output is

英文名，中文名，Peter Wong，John Tam，Sam Hu

because CJK Name is no space between each char, so this solved the big problem for me

worked like a charm ;)

Thankyou @mickmackusa

preg_replace keep the space if start with a-z and end with a-z

3 Answers3